http://www.nrc-cnrc.gc.

ca/irc

Controlling air-borne and structure-borne sound in buildings

NRCC-51340
Quirt, J.D.

September 2009

A version of this document is published in / Une version de ce document se trouve dans:

Inter-Noise 2009, Ottawa, Ontario, August 23-26, 2009, pp. 1-15

The material in this document is covered by the provisions of the Copyright Act, by Canadian laws, policies, regulations and international
agreements. Such provisions serve to identify the information source and, in specific instances, to prohibit reproduction of materials without
written permission. For more information visit http://laws.justice.gc.ca/en/showtdm/cs/C-42
Les renseignements dans ce document sont protgs par la Loi sur le droit d'auteur, par les lois, les politiques et les rglements du Canada et
des accords internationaux. Ces dispositions permettent d'identifier la source de l'information et, dans certains cas, d'interdire la copie de
documents sans permission crite. Pour obtenir de plus amples renseignements : http://lois.justice.gc.ca/fr/showtdm/cs/C-42

Page 1 of 15

Controlling air-borne and structure-borne sound in buildings

J. David Quirt
National Research Council Canada
Ottawa, K1A 0R6, Canada

ABSTRACT
In recent years, the science and engineering for controlling sound transmission in buildings have
shifted from a focus on individual assemblies such as walls or floors, to a focus on performance
of the complete system. Standardized frameworks for calculating the overall transmission
including structure-borne flanking, combined with standardized measurements to characterize
sub-assemblies, have advanced these issues from research concepts to engineering practice in
many countries. From studies of relatively homogeneous and isotropic constructions of concrete
and masonry in the 1990s, the technology is now expanding to include the more complicated
behavior of lightweight framed constructions. These advances in measurement-based
calculations offer the potential for better design based on comprehensive prediction of sound
transmission between units in multifamily buildings. To realize that potential, we still must
overcome several challenges. First, the acoustical prediction tools must be suitable for designers
who integrate the many aspects of building performance. Second, the acoustical metrics must
properly reflect how occupants respond to transmitted sound from both typical airborne sources
and impact sources such as footsteps. These concerns pose major challenges for the next decade
both for research and for implementation.

1. INTRODUCTION
This paper attempts to provide an overview of some key advances in dealing with sound
transmission within buildings. It is naturally limited by the authors personal biases, and hence
focuses on issues from a North American perspective, and deals mainly with experimental results
and experiment-based models used to translate the scientific concepts into engineering practice.
Inevitably it overlaps to some degree with other presentations at this conference, and readers are
directed to the obviously pertinent keynote paper on impact sound sources 1 . To minimize the
overlap with that paper, and with other recent presentations on footstep noise2 , this paper focuses
mainly on transmission of sound from airborne sources, especially in the context of multi-family
residential buildings. To relate the discussion to practical concerns, the paper addresses:
Can we accurately predict transmission to the receiver?
What are the sound transmission paths of concern?
Do available criteria reflect how people react to the transmitted sound?
How can we effectively package the technology for the intended users?

Page 2 of 15

A. Shifting to a new paradigm

Until the last decade (with some notable exceptions 3 ), research on sound transmission between
rooms in buildings has focused mainly on sound transmission through individual assemblies.
This perspective is still evident in North American building codes, which for many decades have
considered only the ratings for the assembly separating adjacent dwellings: Sound Transmission
Class (STC) or Field Sound Transmission Class (FSTC) for airborne sources 4 or Impact
Insulation Class (IIC) for footstep noise 5 .
Airborne
Sound
Source

Transmission
through wall

Airborne
Sound
Source

Flanking Transmission
via ceiling surfaces
Transmission
through wall
Flanking Transmission
via floor surfaces

Separating assembly

Figure 2: Drawings show a cross-section through a building with two adjacent dwellings. Some of the sound from
an airborne source in one unit (represented by red loudspeaker in the drawings, which could represent anything from
a home theatre to people talking loudly) is transmitted to the adjacent unit. The traditional approach (at left) focuses
on only the direct sound transmission through the separating assembly. In reality there are many paths for sound
transmission a few are shown in the right hand drawing - and indirect paths often dominate.

Implicit in this approach is the simplistic assumption (illustrated at left in Figure 2) that sound is
transmitted only through the obvious separating assemblythe separating wall assembly when
the units are side-by-side or the floor/ceiling assembly when units are one above the other. If
there is a problem with the sound insulation, this is ascribed to errors in either design of the
separating assembly or the workmanship of those who built it. Unfortunately, this paradigm is
still predominant among designers and builders in North America.
In reality, the problem is more complex (as illustrated at right in Figure 2)the airborne
sound source excites all the surfaces in the source space. All the surfaces vibrate in response, and
some of this vibration is transmitted across the surfaces abutting the separating assembly,
through the junctions where these surfaces join the separating assembly, and into surfaces of the
adjoining space, where part is radiated as sound. It follows that the sound insulation between
adjacent dwellings is always worse than the sound insulation provided by the obvious separating
assembly. Of course, this has long been recognized in principle (and the fundamental science
was largely explained by Cremer and Heckl 6 decades ago)the problem was to reduce the
complicated calculation process to manageable engineering that yields quantitative estimates.
Occupants of the adjacent space actually hear the combination of sound due to direct
transmission through the separating assembly and any leaks, plus sound due to structure-borne
flanking transmission involving all the other elements coupled to those assemblies. For design or
regulation, the terminology to describe the overall sound transmission including all paths is well
established. ISO ratings 7 such as the Weighted Apparent Sound Reduction Index (Rw) have
been used in many countries for decades, and ASTM has defined the corresponding Apparent
Sound Transmission Class (ASTC), which is used in many examples in this talk. There are other
variants using different normalization or weighting schemes that have arguable advantages, but
this paper uses ASTC as the basic measure of sound insulation for airborne sound.

Page 3 of 15

While measuring the ASTC in a building is quite straightforward, predicting the ASTC due
to the set of transmission paths in a building is quite complex, and requires data on structureborne transmission that is only gradually becoming available.
Most of the remainder of this paper is an overview of experimental results and experimentbased models that have been developed to predict the overall sound insulation between adjacent
spaces in a building. But first, to assess whether the predicted ASTC or Rw is adequate, criteria
must connect the physical performance to the reaction expected from building occupants.
B. Ratings and subjective criteria
For efficient design, we need design criteriaobjectives that quantify acceptable levels of
noise from obvious sources. For the occupants of a building, that includes noise from outdoor
sources such as highways and aircraft, noise from appliances and building services (plumbing,
ventilation, etc.) and noise from neighbours. This talk focuses on noise from neighbors in multifamily residential buildings.
Even with that restricted focus (and decades of refining pertinent regulations and standards)
the criteria seem to be based more on tradition than on substantial scientific studies of human
response. The ISO 717 standard offers 15 metrics for airborne sound insulation between rooms,
27 for insulation of facades, and 6 for impact sound insulation. As Rasmussen has periodically
documented 8 , even within unified Europe this has led to a bewildering array of national
criteria, and many non-European countries have added further variants. One could make a strong
case for the benefit of continuing recent research efforts in this area 9,10 , especially to assess the
most suitable ratings to handle low frequency sound and special sources such as footsteps and
building services (ventilation, plumbing, etc.) to establish a credible foundation for improved
consensus standards. That is clearly one of the key challenges for the next decade.
To maintain a manageable focus, this paper simply presents some existing consensus
criteria for insulation against airborne sound, expressed in terms of the ASTC metric chosen for
this presentation. Because of the wide variation in national approaches to regulation, comparing
specific regulatory limits is not very instructive, but recent schemes for labeling housingto
provide potential buyers or tenants with a market indication of quality of sound insulation
(among other factors) offer a clearer perspective, shown in Figure 3.

Approximate ASTC

65

60

Netherlands, NEN1070

Denmark, DS490
Finland, SFS5907
Sweden, SS25267
Germany, VDI4100
Korea
France, CQCA

55

50

45

Basic

Better

Best

Figure 3: Criteria for enhanced sound insulation between adjacent units in multi-family buildings in acoustic quality
classification systems for several countries, translated to approximate ASTC scale.

Most of these labeling systems have 2 or 3 classes for acoustic comfort better than the
regulatory minimum; some also have lower classes directly connected with national

Page 4 of 15

requirements. The top categories have been grouped here as basic/better/best clusters in
Figure 3. Because the various schemes use different metrics from the set in ISO 717, only an
approximate conversion to ASTC is possible, but that suffices to illustrate the rather small range
of criteria - the Basic quality class requires ASTC in the range from about 50 to 55, and the
Best class requires 60 to 65. The existence of a range in requirements is not surprising given
the different national traditions both for regulations and social expectations. But despite strong
individualism in national expression of the requirements, it appears that there is a fairly clear
consensus on how much sound insulation is good enough to satisfy occupants.
For practical design objectives, the requirements for typical occupants seem fairly clear:
ASTC ~ 52 is good enough to satisfy most of the people, most of the time.
ASTC ~ 65 (maximum of top class range) should provide satisfaction almost always.
From a Canadian perspective, these criteria are quite consistent with the social response
data obtained by Bradley in a survey of 300 pairs of neighbors, living side-by-side in multifamily residential buildings 11 . After obtaining survey responses from each pair of neighbors, the
survey team measured ASTC between the dwellings. As expected there was a range of
responses, but there were clear trends in the mean responses, varying from significant annoyance
when ASTC was under 50 to negligible annoyance (and reporting not hearing sound from the
neighbors) when the ASTC approached 65. These results were broadly consistent with the
criteria proposed above and with the market classification schemes shown in Figure 3.
However, it must be recognized that these criteria are at best fuzzy targets, because many
factors (noisiness of individual neighbors, ambient levels due to building services and intruding
outdoor sound, sensitivity of individual listeners, etc.) ensure that any assessment of social
response versus sound insulation will exhibit significant variance.
Improved measures for the sound insulation should reduce the scatter in these responses,
and would presumably shift the relative acceptability of some types of construction, especially
for those cases where low frequencies dominate, which are problematic according to anecdotal
evidence. Pursuing the refinement of the ratings is worthwhile, especially if clear international
consensus can be established. But for purposes of this paper, the criteria noted above give
reasonable working indications of acceptability in terms of the current metrics.
2. TRANSMISSION IN HEAVY MONOLITHIC CONSTRUCTION
Significant advances in predicting the sound transmission through the complete building system,
including the direct and indirect paths, occurred first for heavy monolithic construction, with
structural elements such as concrete floors and masonry walls. These systems are well-suited to
modeling using statistical energy analysis (SEA) to calculate the transmission - the elements
such as floors or walls can be treated as homogeneous and isotropic, they are lightly damped so
they can reasonably be characterized by reverberant levels, and most energy losses are due to
transfer to adjoining elements. Craik and others advanced this subject from research studies to
text books3. By the mid 1990s SEA was part of widely accepted engineering practice.
This engineering concept was implemented in European standard EN 12354, which was
published in 2000, with parts to address airborne and impact sound transmission between rooms
within buildings and the transmission of outdoor sound into a building. In 2005 the Parts of
EN 12354 were adopted as international standards, as ISO 15712, Building acoustics
Estimation of acoustic performance of buildings from the performance of elements 12 . Although
they are most easily and accurately applied to heavy monolithic structures, these standards also
include extensions to deal with other types of assemblies.

Page 5 of 15

For two adjacent rooms, either side-by-side or one above the other, sound is transmitted
both directly through the separating assembly and via a set of indirect paths involving all the
surfaces connected at each junction common to both rooms. In the simple case, where room
dimensions match, there would be four such junctions, one at each of the four edges of the
separating wall or floor assembly. There is a set of indirect paths for each junction, each path
involving the transfer of energy from a surface in the source room to one in the receiving room.
For heavy monolithic constructions this transfer can be calculated, depending on junction
geometry, and readily established properties of the joined assemblies. For more complex
assemblies, measurement protocols were developed to characterize junction performance 13 .
The practicality of the calculation framework comes from the rather straightforward
extension to deal with the incremental effect of linings added to the basic structural elements.
It is common practice, especially in residential buildings, to add finish surfaces to the basic
structural wall and floor assemblies for example, various multi-layer floor surfaces, or gypsum
board wall and ceiling surfaces that mask both the bare concrete and the building services such
as wiring and pipes. These additional layers can significantly improve the sound attenuation,
both by reducing the transmission of vibration between the lining and the supporting assembly,
and by changing radiation efficiency of the exposed surface. If the lining is treated as simply
changing the sound power flow from the reverberant sound field in the room to the reverberant
vibration in the structural assembly, then as shown conceptually in Figure 4, the practical
calculation combines the basic flow of structure-borne power via the coupled structural elements,
with simple incremental effects due to the linings. Fortunately this approach works well for
heavy monolithic supporting structures.

Figure 4: Transmission combines direct path through separating wall (1) and structure-borne flanking via: wallfloor path (2), floor-wall path (3) and floor-floor path (4), plus corresponding set of paths at other junctions.
Transmission via these paths is altered by addition of linings in the source room and/or receiving room.

The effect of a lining added to a structural base assembly can be determined to first order
by measuring the change in direct sound transmission when the lining is added to a similar base
assembly separating the two rooms of a standard sound transmission laboratory suite. This

Page 6 of 15

process for evaluation of linings was outlined in ISO 15712, and subsequently fleshed out more
completely with a set of reference base assemblies in ISO 140-16 14 . For the flanking paths, this
estimate must be corrected to remove the non-resonant component, and the effect of the lining
depends on the mobility of the base assembly, but the process can provide very good estimates of
the overall performance, especially for heavy concrete or masonry constructions, for which
ISO 15712 estimates should be within a standard deviation of 1.5 dB.
Although extensions to include other types of floor and wall assemblies in the ISO 15712
framework have been investigated, there are significant technical complications that must be
considered for lightweight framed construction 15,16 .
3. TRANSMISSION IN LIGHTWEIGHT FRAMED CONSTRUCTION
Rather than attempt to fit sound transmission for lightweight framed construction into the
framework developed for heavy monolithic systems, research in Canada has focused on
developing an approach customized for performance of typical North American wood-framed
buildings.
A. Concepts for flanking in lightweight constructions
In this approach, developed by Nightingale et al 17 , the power flow via each flanking path is
defined by five transmission factors whose combined effect is characterized by a path transfer
function specific to the type of excitation (airborne or impact) and the construction detail. This
is most simply explained in the context of impact sources. Figure 5 identifies the factors
controlling the transmission of structure-borne sound to the room beside, and the resulting
vibration levels across the floor surface are illustrated in Figure 6, for one position of a standard
tapping machine on a lightweight floor.

1. Power injected by
source depends on
Impedance match

2. Attenuation
with Distance
2

5. Flanking sound
power depends on
radiation impedance

3
3. Junction
Attenuation

4. Structural
Attenuation

Figure 5: Five factors that affect flanking transmission via the floor/wall junction, with an impact source.
Separating
wall

Impact
source

Floor
joists
70

100

Plan view of
floor surface

94
96
98

92 88
86
76
80
82
72
70
78
68
74

70
66
66

Figure 6: Variation across the floor surface of the vibration levels (2kHz band) due to an impact source. The
floor construction has wood joists perpendicular to the separating wall between the two side-by-side rooms.

Page 7 of 15

A general model for such a system must account for all five factors indicated in Figure 5,
for a realistic range of source positions. Clearly the system is anisotropic and highly damped
the vibration field exhibits a strong gradient that is different in the directions parallel and
perpendicular to the joists. In general, this vibration field is a poor approximation of a diffuse
field, which limits the applicability of simple SEA models. Not only do vibration levels vary
strongly across the surface of the structural assembly, but also some added linings (such as floor
toppings) change the attenuation across the structural assembly, with different changes in the
three orthogonal directions pertinent to direct and flanking transmission. Hence, a simple
correction for a given lining (derived from measurement of direct transmission and then used to
correct structure-borne flanking transmission via the supporting structural assemblies) is not
generally applicable for lightweight framed assemblies. The direction of transmission relative to
the framing members becomes an additional parameter needed for accurate prediction.
Essentially the same five factors apply to characterizing the propagation with an airborne
source, as indicated in Figure 7. With an airborne source, the effect of source position is largely
eliminated because there is fairly uniform incident sound power on the surfaces of the room, but
all five factors still affect the sound power reaching the receiving room via the flanking paths as
illustrated in Figure 7 for a subset of the paths at a floor/wall junction.
Changing construction details will alter one or more of the five factors. For example,
linings commonly affect both the attenuation across the underlying structural assemblies and the
power flow to/from the underlying assembly.

Direct Transmission
5

1. Power injected
from airborne
source depends
on impedance

5. Flanking sound power

depends on radiation
impedance

4
2
2. Attenuation
with Distance

3. Junction
Attenuation

4. Structural
Attenuation

Figure 7: Five factors that affect flanking transmission, with an airborne source for the paths involving the floor
surface in the source room. Similar factors apply for all other paths.

Experimental results demonstrating these behaviors, for both airborne and impact sources
driving specific wood framed assemblies, were presented at preceding Inter-Noise conferences 18 .
B. Examples of flanking transmission in lightweight constructions
A few examples to illustrate the effects due to common variations in construction are presented
and discussed here, to provide context for the semi-empirical prediction methods presented
subsequently. The discussion concentrates mainly on one set of base assemblies, but other
systems show comparable trends.
Figure 8 shows a specific set of constructions where a wall separates two side-by-side units;
the wall has gypsum board screwed directly onto one side of the wood stud framing and mounted
on resilient metal channels on the other, and achieves STC 52 in laboratory testing. The floor

Page 8 of 15

assembly has a bare oriented strand board (OSB) floor surface, with its gypsum board ceiling
mounted on resilient channels (STC 55 in laboratory testing).
STC 52
Direct
Transmission

70
Apparent TL for Specific Paths, dB

Direct path through wall

60

Flanking
via subfloor

50

Apparent
STC
44 to 45

40
30

Floor-Floor Path
(Bare floor,Joists parallel wall)

Floor joists parallel

to separating wall
(non-loadbearing wall)

20
10

Alternate
junction
details

0
63

125

250
500
Frequency, Hz

1k

2k

4k

Figure 8: Sound transmission between side-by-side units with simple wood-frame wall and floor assemblies, as
illustrated.

In repeated tests with minor variations of the materials and in the floor/wall junction details, the
overall sound insulation observed between the side-by-side rooms was ASTC 43 to 45.
Measurements of direct transmission through the wall itself showed that its sound transmission
in the complete building system is very similar to laboratory results (STC 52). The difference in
the system performance is due to flanking transmission via the floor assembly, which transmits
far more sound than the separating wall assembly above 250 Hz.
70

STC 52

Apparent TL for Specific Paths, dB

Direct path through wall

60

Direct
Transmission

50

Flanking via
subfloor & joists

Apparent
STC 42

40
30

Floor-Floor Paths (Bare floor)

Joists parallel wall

20

Joists wall

Floor joists
perpendicular to
separating wall
(loadbearing wall)

Joists and continuous

10
0
63

125

250
500
Frequency, Hz

1k

2k

4k

Figure 9: Modifying the wall/floor system of Figure 8 by reorienting the floor joists to run perpendicular to the
separating wall lowers the ASTC for the system.

For the case shown in Figure 9, the measured ASTC was even lower than the ASTC observed
when the joists were parallel to the separating wall (as illustrated in Figure 8). The problem here
is not that the separating wall assembly is transmitting more sound than expectedit is
performing as designedbut that most of the sound energy is able to circumvent the separating
wall as structure-borne flanking transmission. Once again, the system ASTC is much lower than
the STC of the separating assembly because flanking has not been properly considered in the
design.

Page 9 of 15

The systems illustrated in Figures 8 and 9 would result in noise that most neighboring
occupants would find annoying and would complain about. To remedy this, a builders first
impulse would likely be to fix the separating wall assembly by, for example, sealing any
possible leaks and adding a second layer of gypsum board on the side with resilient channels.
The added gypsum board should increase the wall assemblys STC by about 5. Detailed testing
would show that the sound transmission directly through the wall was reduced (i.e. Field STC
increased) as expected, but that the system performance was barely affected and only increased
to ASTC 43 because the dominant sound transmission path (i.e., structure-borne flanking via the
floor) was not dealt with.
In recent years many enhanced products have been introduced, such as wallboard
incorporating constrained-layer damping, or resilient mountings that improve on the traditional
generic resilient metal channels of the walls in Figures 8 and 9. Such products could increase
this basic wall assemblys sound insulation to a rating of STC 60 or more, but the complete
system would still provide only ASTC 43.
To address the problem, one must identify the key sound transmission paths and take
appropriate measures to manage them. As illustrated in Figure 10, since transmission via the
floor is the dominant problem with the floor/wall systems illustrated in Figures 8 and 9, treating
the floor must be part of the solution. But a rational approach to the design must balance
changes to the floor surface with changes to the separating wall, to achieve a cost-effective
system with the desired ASTC performance. If the target were ASTC of at least 50, then a rather
complex and expensive treatment of the floor would be required if using the basic wall illustrated
in Figure 9. A simpler floor treatment could provide the target ASTC if the wall were improved
to STC 57 with an extra layer of gypsum board. With further enhancement of the wall surfaces,
the ASTC could be increased to ~60 when combined with the best floor treatment illustrated in
Figure 10.
Direct
Transmission
Changed flanking
via floor surfaces

Wall
choices

STC
52

STC
57

48

50

Add floor topping:

ASTC with extra layer of OSB
with 25 mm concrete

49

52

with 38 mm concrete
on resilient mat

51

55

Figure 10: With a range of choices for the wall and floor, the builder can look sensibly at cost/performance
tradeoffs for improvements to the elements that affect the dominant paths, which are the separating wall and the
floor surface in this illustration.

Unfortunately, making improvements to the floor and separating wall is not a complete solution,
as other paths may also be significant, and once better floor and wall assemblies have been put in
place, the sound transmission via other paths will become more obvious. Ceilings and sidewalls
also need to be considered as possible paths of sound transmission.

Page 10 of 15

Only at this stage, can acoustical benefits of specific changes be properly weighed and
balanced against their cost to optimize the cost/benefit for the complete system. The examples
above have focused on side-by-side spaces, but a similar set of tradeoffs is involved when one
considers the case where one dwelling is above another.
This highlights the practical need for a more complicated design framework, as discussed in
the next section.
C. Designing for system performance in lightweight constructions
A simplified guide for design of wood-framed buildings was developed 19 , using a tabular
approach to present alternative choices for all the surfaces likely to be significant to the overall
sound transmission between adjacent spaces. The Guide presents single-number ratings for the
transmission of sound from both air-borne and impact sources, for adjacent units that are sideby-side, or one above the other, for a limited set of the most common constructions.
A few examples for airborne sources are presented here to highlight the strengths and
weaknesses of such an approach.
Figure 11 illustrates the situation typically found in apartment buildings. In single-level
apartments, the gypsum board ceiling is normally mounted on resilient channels to enhance the
sound insulation from the apartment above. This also reduces flanking transmission between the
side-by-side units via the ceiling/ceiling path to an insignificant level.

(Ceiling surfaces isolated)

Airborne
Sound
Source

Transmission
through wall
Transmission via
floor surfaces

Ceiling gypsum board on resilient channels

Separating wall
Attachment of gypsum
board on sidewall
Floor surface
No topping (basic)
19-mm OSB
stapled to subfloor
25-mm gypsum concrete
bonded to subfloor
38-mm gypsum concrete
+ resilient mat on subfloor

Basic wall
Better wall
(STC 52)
(STC 57)
Direct or Direct
Resilient
resilient
Apparent STC (ASTC)
43
43
43
48

50

50

49

51

52

51

53

55

Figure 11 and Table 1: Typical sound transmission paths between adjacent one-level apartment units. The sidewalls
abutting the separating wall also transmit sound, but resilient channels supporting the gypsum board ceiling block
transmission via the ceiling/ceiling path. The table presents the apparent STC for the specific separating wall and
floor constructions illustrated, with various treatments of floor and wall surfaces.

From Table 1, the effects of variations in the construction are readily seen. For example, with no
topping added over the basic plywood or OSB floor surface, flanking via the floor surfaces is so
strong that the ASTC between the adjacent units does not rise above 43 no matter what
improvements are made in the separating wall or the sidewalls. Once the floor has been treated,
then the effect of improving the separating wall becomes obvious. With the combination of a
better floor and better separating wall, then the effect of improving the sidewalls also becomes
significant. A paper by Nightingale at this conference 20 addresses this issue in more detail.
In applications where transmission between storeys within a dwelling unit is not a concern
(e.g., row housing), the ceiling is typically screwed directly to the bottom of the joists, as shown
in Figure 12. In such cases, the flanking paths via the ceiling also become significant, and this
reflected in the lower ASTC values in Table 2 for this building design scenario.

Page 11 of 15

Separating wall
Airborne
Sound
Source

Flanking via ceiling surfaces

Transmission
through wall
Flanking via floor surfaces

Attachment of gypsum
board on sidewall
Floor surface
No topping (basic)

Ceiling gypsum board screwed to joists

19-mm OSB
stapled to subfloor
25-mm gypsum concrete
bonded to subfloor
38-mm gypsum concrete
+ resilient mat on
subfloor

Basic wall
(STC 52)

Better wall
(STC 57)

Direct or Direct
Resilient
resilient
Apparent STC (ASTC)
42

43

43

47

48

49

48

49

50

49

51

52

Figure 12 and Table 2: . Typical sound transmission paths between side-by side units in multi-level row housing.
The sidewalls abutting the separating wall also transmit sound. The table presents the ASTC for the specific
separating wall and floor constructions illustrated, with various treatments of floor and wall surfaces.

The corresponding effects when one unit is below another are less dramatic, but still warrant
design consideration. The only significant flanking paths involve the floor surface and the walls
in the room below. Transmission via the wall/wall paths shown in Figure 13 is typically weak
enough so that it can be ignored. The flanking transmission in this case is essentially the same for
all the framing variants tested. The effect of joist orientation (stronger flanking via the walls
supporting the floor joists) averages out if all wall surfaces in the room below are the same,
because the joists are perpendicular to two walls and parallel to the others.
Airborne
Sound
Source
Airborne
Sound
Source

wall-wall
path
floor-wall
path

Direct
Transmission
through floor

Figure 13. Transmission paths between upper and lower units include both direct transmission through the
separating floor and flanking transmission involving the floor and wall assemblies.

Table 3 shows the combined effect of changes to the floor surface, the ceiling and the walls, and
allows one to perform a cost/benefit analysis for different design options. This approach (which
follows the same pattern as that used for side-by-side units) is especially helpful when used with
lightweight floor surfaces.

Page 12 of 15

Walls
in room below

Floor surface

No topping (OSB subfloor)

Basic walls:
All walls with 1 layer of
gypsum board fastened
directly to the studs

19-mm OSB
stapled to subfloor
25-mm gypsum concrete
bonded to subfloor
38-mm gypsum concrete
+ resilient mat on subfloor
No topping (OSB subfloor)

Flanking suppressed:
All walls with 1 layer of
gypsum board supported
on resilient channels

19-mm OSB
stapled to subfloor
25-mm gypsum concrete
bonded to subfloor
38-mm gypsum concrete
+ resilient mat on subfloor

Better ceiling
Worse ceiling
1 layer gypsum board on 2 layers gypsum board on
resilient metal channels
resilient metal channels
spaced 600 mm o.c.
spaced 400 mm o.c.
(STC 59 if no topping)
(STC 51 if no topping)
Apparent STC (ASTC)
49
52
54

59

59

61

63

64

51

59

55

64

62

70

66

74

Table 3. Apparent STC between units (one unit below another) for selected variations of the floor/ceiling assembly
and the wall surfaces in the room below.

Comparison of the ASTC values in Table 3 for a chosen floor topping show that because
flanking transmission via the walls of the room below is comparable to direct transmission
through typical ceilings with resilient channels, expensive solutions to improve the ceiling are
not likely to provide much improvement in the ASTC, unless combined with wall improvements.
Because both the direct transmission path and the significant flanking paths involve the floor
surface, adding extra materials over the bare floor surface is often the most effective way to
improve the sound insulation between units. When all three surfaces (floor, ceiling, and walls
below) are improved, then very good overall performance can be achieved.
A similar set of tables in the Guide present impact (footstep noise) ratings for the same set
of constructions. Thus the simple table-based design guide does provide information on sound
transmission by the complete system, in a form that generalists can use, for a limited set of
practical constructions.
D. Making the design process usable in practice
The tabular approach discussed above does show the effect of changes to all of the surfaces
controlling sound transmissionboth the separating assembly and the key flanking paths (hence
indicating obvious choices)and it also provides ASTC estimates for designers. Because tables
are readily presented in conventional technical documents, distribution of the tabular Guide
provided an effective means to convey concepts to builders and their generalist designers. But
there are some obvious limitations:
Each table (such as Table 1 or Table 2 above) applies to one specific combination of wall
and floor constructions; therefore, many tables were required.
A table can only present a few variants on each of the possible elements such as choices
for floor toppings, or for floor coverings, or for gypsum board type and attachment on
flanking surfaces. This seriously limits the range of options that can be presented.
The tabular approach does not readily support comparison of different designs, or show the
relative significance of the direct and flanking transmission paths in each case.

Page 13 of 15

The obvious means to display more choices for each of the component materialsand to
facilitate a more detailed analytic approachis to implement the calculation framework in
software, linked to a database of sound transmission data for each path, for the matrix of
construction options that have been characterized. For the SEA approach (which is applicable to
heavy monolithic construction as described in Part 2 of this paper) commercial software
packages are available.
A software system is also being developed to implement the approach outlined in Part 3 for
lightweight framed constructions. Such software can easily present a much broader range of
construction options than the tabular approach illustrated in Part 3C. A screen image of the user
interface is shown in Figure 14, to illustrate the potential of such tools to provide acoustical
performance estimates in a form useful for generalists dealing with building design.

Overall Performance:
Apparent STC 51
Apparent IIC 55

3b

2
1
Separating
Partition

Bottom

Top

Front

Back

62

62

--

--

Direct or Flanking - STC

3a

57

53

58

Direct or Flanking - IIC

--

55
53

--

Impact not applicable

for selected junction

Figure 14: Example of user interface to illustrate how software can facilitate the display of sound transmission
estimates for the set of transmission paths between adjacent spaces, to guide design decisions and estimate system
performance. Parts of the interface include: (1) buttons to select between the separating assembly or each of the
four flanking junctions at its edges, (2) drop down menus to select details of framing and other components
affecting transmission via the selected junction, (3a) calculated sound transmission ratings for each set of paths,
(3b) calculated overall sound insulation estimate.

An interface like that shown in Figure 14 can provide an interactive framework where the
designer can explore changes in the building assemblies and materials to balance the sound
transmission via the separating assembly and the set of flanking paths for the four junctions, in
addition to giving ratings of the overall sound insulation.
These acoustical performance estimates provide the acoustical part of the information
matrix needed by a design team for rational tradeoffs between the effect of specific changes in
the building elements on the noise control, versus their impact on cost and other design
objectives for building performance, such as fire resistance, structural capacity and energy use.

Page 14 of 15

The balancing of many performance requirements is central to efficient design, and at the
heart of the integrated design process central to modern green building schemes. Providing
tools to support the acoustics part of satisfying the design requirements is essential to having
acoustical performance effectively integrated into such schemes.
4. SUMMARY
The engineering framework to deal with sound transmission between neighboring units in
complete buildingsboth experimental techniques to characterize subsystems and calculation
methods to turn the experimental data into estimates of sound insulationhas largely been
developed. Design tools to make the knowledge readily accessible to design generalists are
rapidly becoming available. This is enabling a paradigm shift from the traditional simplistic
focus on the separating assembly, to properly evaluating performance of the complete building
system.
ACKNOWLEDGMENTS
The author gratefully acknowledges the contributions of colleagues in the Acoustics Group of
the Institute for Research in Construction at NRC, especially Trevor Nightingale, Alf Warnock,
and Robin Halliwell. Not only did they share in the development of key concepts reported here,
but also, they contributed steadily to my education in building acoustics through decades of
collaboration. I also acknowledge the repeated stretching of my perspectives provided by many
colleagues in the working groups of ISO/TC43/SC2. Although I am the nominal author, this
paper is truly a summary of the work of many others.
REFERENCES
1

Jin Yong Jeon, Building impact sound sources and ratings, Inter-Noise 2009, Ottawa, August 2009
T.R.T. Nightingale, Controlling impact noise in wood-frame multi-unit buildings, International Congress on
Sound and Vibration, Daejon, Korea, July 2008
3
R.J.M. Craik, Sound transmission through buildings using Statistical Energy Analysis, Gower Publishing Ltd.,
1996 (and extensive references presented there)
4
ASTM E336-05, Standard Test Method for Measurement of Airborne Sound Insulation in Buildings and ASTM
E413, Classification for Rating Sound Insulation, ASTM International, West Conshohocken, PA.
5
ASTM E1007, Test Method for Field Measurement of Tapping Machine Impact Sound Transmission Through
Floor-ceiling Assemblies and Associated Support Structures, and ASTM E989 Classification for
Determination of Impact Insulation Class (IIC), ASTM International, West Conshohocken, PA.
6
L. Cremer, M. Heckl, and E.E. Ungar, Structure-borne sound, 2nd edition, Springer-Verlag, New York, 1988.
7
ISO 717, AcousticsRating of sound insulation in buildings and of building elementsPart 1: Airborne Sound
Insulation, Part 2: Impact sound insulation International Organization for Standardization, Geneva. ISO rating
Rw is approximately the same as ASTC, and Lnw is approximately 110- AIIC.
8
Birgit Rasmussen, Sound insulation between dwellings Update on classification schemes and building
regulations in Europe, CIB W051 meeting in Bors Sweden, May 2008.
9
H. K. Park, J. S. Bradley, and B. N. Gover, Evaluating airborne sound insulation in terms of speech
intelligibility, JASA, Vol 123, pp 1458-1471, March 2008
10
Jin Yong Jeon, Pyoung Jik Lee, Jae Ho Kim, and Seung Yup Yoo, Subjective evaluation of heavy-weight floor
impact sounds in relation to spatial characteristics, JASA Vol. 125, 2987-2994, May 2009.
11
J.S.Bradley, Deriving Acceptable Values for Party Wall Sound Insulation from Survey Results, Proceedings of
Inter-Noise 2001, The Hague, Netherlands, August 2001.
12
ISO 15712 Building acoustics estimation of acoustic performance of buildings from the performance of
elements - Part 1: Airborne sound insulation between rooms, Part 2: Impact sound insulation between rooms,
Part 3: Airborne Sound Insulation against outdoor noise, Part 4: Transmission of indoor sound to outside , 2005.
13
ISO 10848, Acoustics- laboratory measurements of flanking transmission of airborne and impact noise between
adjoining rooms Part 1: Frame document, International Organization for Standardization, Geneva, 1998.
2

Page 15 of 15

14

ISO 140-16, Acoustics Measurement of sound insulation in buildings and of building elements Part 16:
Laboratory measurement of the sound reduction index improvement by additional lining, International
Organization for Standardization, Geneva, 2007
15
Stefan Schoenwald, Flanking sound transmission through lightweight framed double-leaf walls, Ph.D. Thesis,
Technische Universiteit Eindhoven, Eindhoven, The Netherlands, 2008.
16
E. Gerretson, Some aspects to improve sound insulation prediction models for lightweight elements,
Proceedings of Inter-Noise 2007, Istanbul, Turkey, August 2007.
17
T.R.T. Nightingale, R.E. Halliwell, J.D. Quirt, Vibration response of floors and the effectiveness of toppings to
control flanking transmission, Proceedings of Inter-Noise 2002 Dearborn, USA, August 2002
18
J.D. Quirt and T.R.T. Nightingale, On a semi-empirical approach to predicting sound insulation in lightweight
framed construction, Proceedings of Inter-Noise 2007, Istanbul, Turkey, August 2007
19
J.D. Quirt, T.R.T Nightingale, and F. King, Guide for Sound Insulation in Wood Frame Construction, RR-219,
NRC-IRC, Canada, March 2006.
20
T.R.T. Nightingale, B. Zeitler, S. Schoenwald, F. King, A hierarchy of flanking transmission paths in lightweight
wood frame construction, Proceedings of Inter-Noise 2009, Ottawa, August 2009